jkrishnavs / openmpgraphalgorithms

1.Introduction

Realization of popular Graph Algorithms using C and OpenMP. The base algorithms are taken from GreenMarl[1] example algorithms. Apart from these we have also created a few utility graph algorithms

2. Algorithm Details

i. communities
ii. conduct
iii. pagerank
iv. sssp
v. triangle_counting
Utility Algorithms
i. addEdgeWeights: Add random edge weights to an unweighted directed graph to output a weighted directed graph. maxLength(Default value 100) and random seed (Default value 0) can be given as user inputs.
ii. preprocess: Preprocess the input graph to generate an isomer optimized for cache performance.
iii. graphequivalence: given tow isomers and a vertex map verifies the equivalance of the two graphs.
iv. graphprop: Extracts the graph properties from the input graph.
v. train*#*.c: Some random graph kernels.

3. Folder Structure

inputGraphs: a set of real world and generated graphs in EDGE_LIST[1] format. We only handle EDGE_LIST format.

4. Input Graphs

The input graphs are either downloaded from SNAP data set[2] or RMAT graphs generated using SNAP[3] or random graphs generated using GreenMarl graph gen program[1].

***** skipped due to size **** Please contact for set of input graphs.

SNAP data set graphs:

1. amazon.edge
2. higgs.edge
3. patent.edge
4. pokec.edge
5. pokecr.edge
6. rhiggs.edge
7. roadNet-CA.edge
8. rpatent.edge
9. web-BerkStan.edge
10. wiki-topcats.edge

RMAT Graphs:

1. rmatv1e5e1e7.edge
2. rmatv7e5e5e6.edge
3. rrmatv1e5e1e7.edge
4. rrmatv7e5e5e6.edge

Random Graphs:

1. rm1e6u8e6.edge
2. rm1e6u35e6.edge

4.Licence

See GMLICENCE for GreenMarl Licence. See LICENCE.md for our licence.

5. Compilation

Run ./configure.sh before running make.

make modes: i. make ii. make intermediate iii: make debug

To change chunk size make SET_CHUNKSIZE=

Place Holder header: /include/energylib.h

You can replace the function in energylib to track power and energy consumption of the kernel. please see [4] for more details.

6.Additional Dependencies

1. SPRNG: Scalable parallel Random Number Generator: http://www.sprng.org Download and add to the home folder

7.References

[1] https://github.com/stanford-ppl/Green-Marl
[2] https://snap.stanford.edu/data/
[3] SNAP: A General-Purpose Network Analysis and Graph-Mining Library
[4] https://github.com/jkrishnavs/Energymonitorlibrary

8. Contact

1. Jyothi Krishna V S, IIT Madras ([email protected])
2. Rupesh Nasre, IIT Madras